SMARTLOCK: IOT-BASED LOCKING SYSTEM WITH FACE RECOGNITION

SMARTLOCK: IOT-BASED LOCKING SYSTEM WITH FACE RECOGNITION

Anđela Jeftović

In an era characterized by rapid technological advancement and increasingly demanding lifestyles, there is a growing need to modernize approaches to security in residential, commercial, and institutional settings. This thesis addresses this challenge by proposing the design and development of a smart lock system. The system integrates multiple methods of access, including PIN authentication, facial recognition, RFID cards, and a mobile application. The hardware implementation includes components such as the NodeMCU microcontroller, IR and RFID sensors, and the ESP32-CAM module for image capture and recognition. The backend is developed in Python, while a Flutter-based mobile application enables users to monitor lock status, review access logs, and manage system settings in real time. Testing confirmed that the system reliably authenticates users across the different access methods and provides accurate monitoring through the application. The results indicate that multi-modal authentication systems can offer a more robust solution compared to traditional locking mechanisms. This project contributes to the growing field of IoT-based smart home and office security, with potential for future improvements such as cloud integration, advanced encryption, and scalability to larger environments.

Smart Lock, Internet of Things (IoT), Access Control, Face Recognition, RFID authentication, PIN code authentication, mobile application, Flutter, ESP32-CAM, NodeMCU, microcontroller, FastAPI, ResNet50, MQTT protocol.